Security Audit

bv

github.com/Mrc220/agent_flywheel_clawdbot_skills_and_integrations

AI SkillCommit c7bd8e0f6900

TRUSTED

Scanned about 2 months ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

bv received a trust score of 83/100, placing it in the Mostly Trusted category. This skill has passed most security checks with only minor considerations noted.

SkillShield's automated analysis identified 2 findings: 0 critical, 1 high, 1 medium, and 0 low severity. Key findings include Potential Command Injection via unsanitized filename in `bv --export-graph`, Potential Prompt Injection via `beads.jsonl` content in `bv` output.

The analysis covered 4 layers: Manifest Analysis, Static Code Analysis, Dependency Graph, LLM Behavioral Safety. All layers scored 70 or above, reflecting consistent security practices.

Last analyzed on February 11, 2026 (commit c7bd8e0f). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

100%

Static Code Analysis

100%

Dependency Graph

100%

LLM Behavioral Safety

78%

Behavioral Risk Signals

Filesystem Write

1 finding

Shell Execution

2 findings

Dynamic Code

2 findings

Security Findings2

Severity	Finding	Layer	Location
HIGH	Potential Command Injection via unsanitized filename in `bv --export-graph` The `bv --export-graph <file.html>` command allows specifying an output filename. If an AI agent constructs this command using untrusted user input for `<file.html>` without proper sanitization, a malicious user could inject shell commands (e.g., `bv --export-graph "malicious.html; rm -rf /"`). This could lead to arbitrary command execution on the agent's host. AI agents should strictly sanitize or validate any user-provided input for filenames passed to `bv --export-graph`. Only allow alphanumeric characters, hyphens, underscores, and a single dot for the extension. Alternatively, use a fixed, agent-controlled temporary directory and filename.	LLM	SKILL.md:172
MEDIUM	Potential Prompt Injection via `beads.jsonl` content in `bv` output The `bv` tool processes user-controlled content from `.beads/beads.jsonl` (e.g., bead titles or descriptions). This content can appear in the `reason` field of `bv --robot-triage` output (e.g., in `recommendations`). If malicious instructions are embedded within the `beads.jsonl` data, and an AI agent then consumes and presents this `reason` field directly to its host LLM without sanitization, it could lead to prompt injection, manipulating the LLM's behavior. AI agents consuming `bv` output should sanitize or filter any text fields (like `reason`) before presenting them to the host LLM. The `bv` tool itself could also offer an option to strip or escape potentially malicious LLM instructions from its text outputs.	LLM	SKILL.md:192

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/3523aadb2d42b385.svg)](https://skillshield.io/report/3523aadb2d42b385)